Researcher, Pretraining Safety
OpenAI(3 months ago)
About this role
The Researcher, Pretraining Safety at OpenAI is responsible for developing techniques to evaluate and reduce unsafe behaviors in base models during pre-training. This role involves designing architectures and training setups to prioritize safety, as well as collaborating with cross-functional teams to create scalable safety foundations. Key responsibilities include implementing safety-oriented metrics, refining data curation strategies, and exploring novel model configurations that enhance controllability while mitigating risks.
Required Skills
- Safety Evaluation
- Risk Reduction
- Architecture Design
- Model Strengthening
- Data Curation
- Novel Loss Functions
- Cross-Functional Collaboration
- Statistical Reasoning
- Research Workflows
- Experiment Iteration
+5 more
About OpenAI
openai.comThe company specializes in providing innovative solutions to enhance online experiences for businesses and consumers. It offers a range of products and services designed to streamline operations, improve customer engagement, and drive growth in the digital landscape. With a focus on user experience and cutting-edge technology, the company aims to empower businesses to thrive in a rapidly evolving market.
Apply instantly with AI
Let ApplyBlast auto-apply to jobs like this for you. Save hours on applications and land your dream job faster.
More jobs at OpenAI
Similar Jobs
Research Scientist, Pretraining, Gemma
DeepMind(1 month ago)
Tech Lead - Pretraining Team, Wayve Foundation Model
Wayve(6 months ago)
Member of Technical Staff, Video Generation - Audio
xAI(1 month ago)
Research Engineer / Research Scientist, Vision
Anthropic(18 days ago)
Research Scientist – LTX Model Quality
Lightricks(7 days ago)
Member of Technical Staff, Pretraining evaluations
Cohere(1 month ago)